Speech Dereverberation Using Fully Convolutional Networks
This work addresses speech enhancement for applications like hearing aids or communication systems, but it is incremental as it adapts existing image processing networks to a known problem.
The paper tackled speech dereverberation with a single microphone by applying fully convolutional networks to STFT images, finding that their method outperformed competing methods in most cases on the REVERB challenge data.
Speech derverberation using a single microphone is addressed in this paper. Motivated by the recent success of the fully convolutional networks (FCN) in many image processing applications, we investigate their applicability to enhance the speech signal represented by short-time Fourier transform (STFT) images. We present two variations: a "U-Net" which is an encoder-decoder network with skip connections and a generative adversarial network (GAN) with U-Net as generator, which yields a more intuitive cost function for training. To evaluate our method we used the data from the REVERB challenge, and compared our results to other methods under the same conditions. We have found that our method outperforms the competing methods in most cases.